[awade, Craig]

As a side note to this diagnostic oscilloscope probing I noticed the other day that there were 50 MHz oscillations at TP14 and TP15 in the fast path of both field boxes. These points are probing the two active LP filters in the fast stage, an OP27 and AD797 respectively.  Values are as follows:

North (2009:005) 21.6 mVpp @ TP14, 203 mVpp @ TP15, 42 mVpp @ TP16, too small to see at subsequent stages.

South (2009:007) 32 mVpp @TP14, 189 mVpp @ TP15, 40 mVpp @ TP16, too small to see at subsequent stages.

I'm guessing the oscillations are a stability artifact of the OP27 stage one circuit (U8) which is then amplified by the stage two AD797 (U9). After this point in the path the 50 MHz is attenuated by the subsequent stages as they are essentially in buffering G = 1 mode, I'm guessing it sits outside their bandwidth. The size isn't so small but its at very high frequency and quickly filtered out by the circuit: so maybe its not really a problem.  

When I zoomed out a little in the time scale and adjusted the trigger threshold up it seemed like some of these 50 MHz oscillations were being excited by glitches. In the time domain these ring down to the stead state levels listed above.

On another thing, I'm not sure why there are two different types of op amps here.  I guess I find it strange that the OP27 is first, the AD797 spec sheet seems to suggest it has a lower noise than the OP27. Maybe its because the GBW of the OP27 is higher? 

Anyway, just a note for future reference. I have no conspiracy theories that this is an actual issue for us, yet.